Concordance of cancer drug therapy information derived from routinely collected hospital admissions data and the Systemic Anti-Cancer Therapy (SACT) dataset, for older women diagnosed with early invasive breast cancer in England

1 Introduction

National guidelines for women diagnosed with early invasive breast cancer (EIBC) recommend the use of chemotherapy (in addition to surgery), along with targeted therapies where tumour and patient characteristics suggest those treatments would improve survival outcomes. For women diagnosed with human epidermal growth receptor 2 (HER2)-positive EIBC, the targeted therapy trastuzumab is recommended in combination with chemotherapy [1] . The evidence underlying such recommendations is primarily from clinical trials in relatively fit, selected age cohorts which may be limited in their generalisability to older “real world” patient populations [23] . Consequently, it is desirable to be able to evaluate the uptake of oncological treatments, and subsequent outcomes, using national data sources to understand the risks and benefits of treatment outside of a trial setting [4] . Such evaluations depend on the data sources containing accurate and complete information about cancer drug therapy (CDT).

Patient-level data on aspects of breast cancer care are routinely collected in hospitals and mandatorily submitted to national organisations, as part of the care and support of patients with cancer. The Systemic Anti-Cancer Therapy (SACT) dataset collects patient and tumour-level data on CDT (such as chemotherapy and targeted therapy) delivered within secondary and tertiary care settings [5] . Previous publications using the SACT dataset have highlighted incomplete data capture and hospital-level variation in data returns and quality [56] . Studies of patients with lung cancer and colon cancer have compared chemotherapy recorded in the SACT dataset with information in the Hospital Episode Statistics Admitted Patient Care (HES-APC) dataset [78] . These identified that the recording of chemotherapy cycles in the SACT dataset was incomplete, with additional cycles identified in HES-APC and differences in data capture according to patient age and fitness, indicating that both data sources should be used to derive information about chemotherapy. For breast cancer, most CDT treatment is delivered as day case admissions in the secondary care setting, therefore HES-APC may provide an additional data source for identifying CDT use [9] .

Improvement in cancer treatment outcomes requires the translation of recommendations on optimal treatment into delivery of those drugs to patients but there is evidence of considerable variation in this practice [1011] . One aspect of the verification process of what happens in routine care is to examine complete, reliable information on CDT prescription at a patient-level. This has traditionally been a difficult task. The introduction of SACT in England greatly improved the quantity and quality of CDT information available nationally but there remain some gaps. NHS trusts with lower levels of SACT data returns require targeted approaches supported by data derived from other sources to ensure poor data returns do not mask deficiencies in care. Similarly higher levels of variation in cancer care including receipt of CDT are reported for older patients [101213] . SACT alone does not currently meet all these data needs.

The aim of this study was to evaluate the consistency of CDT information recorded within SACT and HES-APC for a cohort of women aged 50 years and over newly-diagnosed with EIBC in England from 2014 to 2019. The rationale for the study was to determine the value of HES-APC in identifying CDT use, and whether it could provide information that complements the SACT dataset.